From 7e01d092a23168b075c3bec27cf7caed7cfe544b Mon Sep 17 00:00:00 2001
From: Ganlin Zhao <ganlinzhao@hotmail.com>
Date: Tue, 21 Feb 2023 17:24:09 +0800
Subject: [PATCH] optimize repeat scan

---
 source/libs/executor/src/executorimpl.c |  5 +++--
 source/libs/executor/src/scanoperator.c | 11 ++++++++---
 source/libs/function/src/builtins.c     |  3 ++-
 source/libs/function/src/builtinsimpl.c |  2 +-
 4 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/source/libs/executor/src/executorimpl.c b/source/libs/executor/src/executorimpl.c
index 2f76a3227b..b02c9c1d57 100644
--- a/source/libs/executor/src/executorimpl.c
+++ b/source/libs/executor/src/executorimpl.c
@@ -395,19 +395,20 @@ void applyAggFunctionOnPartialTuples(SExecTaskInfo* taskInfo, SqlFunctionCtx* pC
   }
 }
 
-static void doSetInputDataBlockInfo(SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t order) {
+static void doSetInputDataBlockInfo(SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t order, int32_t scanFlag) {
   SqlFunctionCtx* pCtx = pExprSup->pCtx;
   for (int32_t i = 0; i < pExprSup->numOfExprs; ++i) {
     pCtx[i].order = order;
     pCtx[i].input.numOfRows = pBlock->info.rows;
     setBlockSMAInfo(&pCtx[i], &pExprSup->pExprInfo[i], pBlock);
     pCtx[i].pSrcBlock = pBlock;
+    pCtx[i].scanFlag = scanFlag;
   }
 }
 
 void setInputDataBlock(SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t order, int32_t scanFlag, bool createDummyCol) {
   if (pBlock->pBlockAgg != NULL) {
-    doSetInputDataBlockInfo(pExprSup, pBlock, order);
+    doSetInputDataBlockInfo(pExprSup, pBlock, order, scanFlag);
   } else {
     doSetInputDataBlock(pExprSup, pBlock, order, scanFlag, createDummyCol);
   }
diff --git a/source/libs/executor/src/scanoperator.c b/source/libs/executor/src/scanoperator.c
index 8528985e05..097d6da63a 100644
--- a/source/libs/executor/src/scanoperator.c
+++ b/source/libs/executor/src/scanoperator.c
@@ -699,7 +699,8 @@ static SSDataBlock* doGroupedTableScan(SOperatorInfo* pOperator) {
 
     if (pTableScanInfo->scanTimes < pTableScanInfo->scanInfo.numOfAsc) {
       setTaskStatus(pTaskInfo, TASK_NOT_COMPLETED);
-      pTableScanInfo->base.scanFlag = REPEAT_SCAN;
+      pTableScanInfo->base.scanFlag = MAIN_SCAN;
+      pTableScanInfo->base.dataBlockLoadFlag = FUNC_DATA_REQUIRED_DATA_LOAD;
       qDebug("start to repeat ascending order scan data blocks due to query func required, %s", GET_TASKID(pTaskInfo));
 
       // do prepare for the next round table scan operation
@@ -725,7 +726,7 @@ static SSDataBlock* doGroupedTableScan(SOperatorInfo* pOperator) {
 
       if (pTableScanInfo->scanTimes < total) {
         setTaskStatus(pTaskInfo, TASK_NOT_COMPLETED);
-        pTableScanInfo->base.scanFlag = REPEAT_SCAN;
+        pTableScanInfo->base.scanFlag = MAIN_SCAN;
 
         qDebug("%s start to repeat descending order scan data blocks", GET_TASKID(pTaskInfo));
         tsdbReaderReset(pTableScanInfo->base.dataReader, &pTableScanInfo->base.cond);
@@ -878,7 +879,11 @@ SOperatorInfo* createTableScanOperatorInfo(STableScanPhysiNode* pTableScanNode,
 
   pInfo->scanInfo = (SScanInfo){.numOfAsc = pTableScanNode->scanSeq[0], .numOfDesc = pTableScanNode->scanSeq[1]};
 
-  pInfo->base.scanFlag = MAIN_SCAN;
+  if (pInfo->scanInfo.numOfAsc > 1) {
+    pInfo->base.scanFlag = REPEAT_SCAN;
+  } else {
+    pInfo->base.scanFlag = MAIN_SCAN;
+  }
   pInfo->base.pdInfo.interval = extractIntervalInfo(pTableScanNode);
   pInfo->base.readHandle = *readHandle;
   pInfo->base.dataBlockLoadFlag = pTableScanNode->dataRequired;
diff --git a/source/libs/function/src/builtins.c b/source/libs/function/src/builtins.c
index 755f103c44..8902bd72cf 100644
--- a/source/libs/function/src/builtins.c
+++ b/source/libs/function/src/builtins.c
@@ -2276,8 +2276,9 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
   {
     .name = "percentile",
     .type = FUNCTION_TYPE_PERCENTILE,
-    .classification = FUNC_MGT_AGG_FUNC | FUNC_MGT_REPEAT_SCAN_FUNC | FUNC_MGT_FORBID_STREAM_FUNC,
+    .classification = FUNC_MGT_AGG_FUNC | FUNC_MGT_REPEAT_SCAN_FUNC | FUNC_MGT_SPECIAL_DATA_REQUIRED | FUNC_MGT_FORBID_STREAM_FUNC,
     .translateFunc = translatePercentile,
+    .dataRequiredFunc = statisDataRequired,
     .getEnvFunc   = getPercentileFuncEnv,
     .initFunc     = percentileFunctionSetup,
     .processFunc  = percentileFunction,
diff --git a/source/libs/function/src/builtinsimpl.c b/source/libs/function/src/builtinsimpl.c
index 59aa2f1a92..bf4ca7d01e 100644
--- a/source/libs/function/src/builtinsimpl.c
+++ b/source/libs/function/src/builtinsimpl.c
@@ -1578,7 +1578,7 @@ int32_t percentileFunction(SqlFunctionCtx* pCtx) {
   int32_t          type = pCol->info.type;
 
   SPercentileInfo* pInfo = GET_ROWCELL_INTERBUF(pResInfo);
-  if (pCtx->scanFlag == REPEAT_SCAN && pInfo->stage == 0) {
+  if (pCtx->scanFlag == MAIN_SCAN && pInfo->stage == 0) {
     pInfo->stage += 1;
 
     // all data are null, set it completed